# Replication code and data for: Global projections of heat exposure of older adults

# Replication code and data for: Global  projections of heat exposure of older adults
By Giacomo Falchetta, Enrica De Cian, Ian Sue Wing and Deborah Carr
Nature Communications. DOI: 10.1038/s41467-024-47197-5

An output data file, containing grid-cell level counts of people by age group, of climate hazard indicators, and of heat exposure metrics for both the historical climate and current demography and for future scenarios and projections is contained in the repository ("aging_climate_output_data.csv"). 

Software requirements:
- R v4.3+: https://cran.r-project.org/bin/windows/base/
- RStudio: v2023.06.0+: https://posit.co/download/rstudio-desktop/
- Package dependencies: raster, sf, tidyverse, rasterVis, rgdal, maptools, pbapply, terra, knitr, kableExtra, modelsummary, openxlsx, xtable, ggforce, maptools, weights, spatstat, rworldmap, scales, patchwork, stars, viridis, devtools, stargazer, readxl, nominatimlite, urbnmapr

To replicate the analysis:
- Clone the replication code repository from https://github.com/giacfalk/aging_climate
- Download input data from this Zenodo data repository
- Download all the 1km age and gender-stratified global population counts rasters from the following WorldPop page https://hub.worldpop.org/geodata/summary?id=24798 and put them in a subdirectory of the working directory called "AGEPOP"
- Run the "project_pop.R" script to generate gridded age-stratified population data for each SSP scenario
- Run the "compare_pop_projections.R" file to compare the generated gridded age-stratified population data with an array of pre-existing sources from different countries and produce a summary comparison table (NOTE: before running the script, decompress the "new_comparison_data.zip" folder into the working directory)
- Run "projections_exposure_m.R" to quantify heat exposure and generate the figures and tables reported in the paper

To process the data and run succesfully, the script requires a computer with at least 32GB RAM. The running time varies based on CPU characteristics, but a runtime of at least 2 hours should be expected to generate all the output data, figures, and tables. All output files are saved in the working directory.

___

This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.